Prediction of CV risk
As cardiovascular (CV) disease corresponds to the most common cause of death in the United States with estimates exceeding one million deaths annually [
1], estimates of individual and population-based CV risk are of paramount importance. CV risk prediction formulae and tables are decision tools that allow the identification of patients at high risk of CV disease. These tools allow early interventions by providers to recommend lifestyle modification or drugs to control modifiable CV risk factors, including hypertension, diabetes, smoking, dyslipidemia and obesity.
Several CV risk prediction formulae are used in clinical practice worldwide. In the United States, the modified Framingham Risk Score (FRS) is the most commonly used tool [
2], and has been adapted for use in diverse populations in other parts of the world. Other tools include the Prospective Cardiovascular Munster Heart Study (PROCAM) [
3], the Systematic Coronary Risk Evaluation system (SCORE) [
4], United Kingdom Prospective Diabetes Study (UKPDS) [
5] tool for diabetics, the Reynolds Risk Score [
6,
7] and more recently, one which includes obesity as a variable (NHANES) [
8].
The variables included in the FRS include age, sex, smoking status, diabetes status, cholesterol, and blood pressure values. These variables are routinely available in patients receiving medical care, particularly in a primary care setting, as routine screening for hypertension, smoking status, dyslipidemia, and fasting hyperglycemia are part of normative preventative health measures [
9]. With such information clinicians could either use gender-specific risk score tables in assigning points that can translate into a given 10-year CV risk, or use electronic or web-based risk calculators
http://www.framinghamheartstudy.org/risk/coronary.html to calculate such risks. The purpose of risk stratification is to identify and treat patients that may be at higher long-term CV risk in a simple and cost-efficient manner. This risk tool has an acceptable area under the receiver operating curve of roughly 75% [
10]. Most of these tools, particularly the FRS have been validated in many different populations and ethnic groups and recalibrated appropriately [
11‐
14], making it a well-known risk index that allows comparison of risks across different population groups. However, there are distinct ethnic populations, particularly in those with higher prevalences of metabolic syndrome where recalibration is often challenging and its applicability may be limited [
11,
13‐
15].
The advent of other markers of CV disease, including high sensitivity C-reactive protein (HS-CRP), homocysteine, lipoprotein (a), or coronary calcification scores [
16‐
18], have been shown to predict incident coronary disease but they add little prognostic value to standard risk formulae and their incorporation into present clinical practice has been challenging. High sensitivity CRP is associated with an increased risk of future cardiovascular disease, diabetes and even hypertension [
16,
19,
20]. Yet studies have shown a minimal incremental value of adding new biomarkers to existing prediction models in a recent study by the Framingham group [
21]. Interestingly, there is emerging evidence in potentially using CRP, particularly using the Reynolds Risk Score, to re-classify intermediate risk subjects into a high risk category for potential interventions [
6,
7,
22]. In addition, detection of subclinical coronary artery disease with screening tests like CT to measure coronary calcium has been a topic of a recent debate. The AHA Consensus document which outlined the likely benefit of using this modality to risk-stratify intermediate risk patients to either high or low categories depending on the score [
23]. This group did not recommend its use in either low or high-risk patients. The utility of biomarkers for the detection of subclinical coronary disease for risk stratification may be limited in individuals believed to be at high CV risk, who for some reason have an FRS that is not too high, but these biomarkers are possibly useful in those with intermediate risk.